Using AI for Generating Art in Your Solopreneur Business

Last year, I explored the potential of AI for solopreneurs, with some very mixed results. It lead me to looking into AI for generating art – and what an interesting rabbit hole that turned out to be!

But, before I get into my journey with AI generated art, let’s start at the beginning – for those of you who don’t know what I’m talking about with “AI art”…

What the heck is “AI Art” anyway?

AI – or artificial intelligence – simply refers to computer algorithms, or code, that has been designed in such a way that it can learn how to do things on its own. AI art, then, is code that has been designed learn to create different images.

The code has been specifically trained on millions of images, so that it has come to understand what various objects are (eg animals, people, landscapes, etc), what types of images there are (eg paintings, drawings, photographs, 3D), and even certain styles within those image types (eg impressionist art, renaissance art).

So when given a text prompt, such as “a painting of a dog in an impressionist style” it knows what you mean, and can give you an image which fits those parameters.

Of course, its all a bit more complicated than that, but that’s the basic gist of it.

My early exploration of AI Art

I started out by trying a number of free online AI art generators, including Craiyon, Stable Diffusion Web, and Deep AI – and wasn’t terribly impressed with the results!

These were the images from my first foray into AI art – using the text prompt “painting of a cute robot writing a book”

I even ended up with monstrosities like this, when trying to generate images of pretty fairies:

YIKES!

To say I was frustrated was an understatement – in fact, I nearly wrote off the whole “AI art thing” altogether!

But I could see other people getting amazing results with AI art, so I decided to dig a little deeper….

The learning curve with AI Art

What I uncovered was this – that while creating AI art seems to be pretty simple – just enter in a text description and press a button – generating really good images, is anything but!

There is quite a learning curve involved in figuring out what makes a good text prompt. In fact there’s a term – “prompt engineering”- and a number of online communities and websites devoted to it.

So I delved even deeper, installing a free, open source AI art code-base, called Stable Diffusion, on my computer. It was a bit technical stretch for me, but I managed it thanks to some great YouTube videos. (These days, thankfully its a bit easier with the likes of Automatic111 and InvokeAI!).

I spent many, many hours exploring all the in and outs of Stable Diffusion, experimenting with text prompts, negative text prompts, various settings and community created models. And I started to actually create some beautiful images, like these:

And once I could see what was possible, using AI for generating art started to become a real passion of mine. As someone who is very creative, but artistically-challenged, its given me a whole new way to realise all those weird and wonderful visual ideas in my head.

I love to challenge myself, and the AI, pushing the boundaries to see just what we can do together.

And its a hell of a lot of fun to see what comes out of it!

Using AI for Generating Business Images

Before you start thinking that the title of this blog post is a little misleading, I am getting to how you can use AI art in your solo entrepreneur endeavours (I promise!)

To be honest, up until now I’ve mostly been using it for personal stuff, such as backgrounds for inspiration quotes/messages on my FaceBook profile.

But the potential for business use is HUGE!

The days where we budget-constrained solopreneurs are limited to free or cheap stock images for our visuals are over!

With AI art tools, we now create totally unique images for anything – social media, websites, blogs (the header image for this post is AI generated), sales funnels, marketing/promotional materials, ads, book covers, lead magnets. Anything you can think of, is probably possible.

Instead of spending time searching through stock photo sites looking for that perfect image, we can just create it ourselves!

And we can more easily create great-looking images which match our brand – not just colours, but the actual look and feel of our brand.

Its particularly useful if you’re in a field where getting relevant images from your typical stock photo sites may be difficult (eg esoteric fields like energy healers and angel guides)

The Downsides of AI Art

It would be remiss of me to talk up the benefits of AI art, without also mentioning the downsides.

The first, is that images produced by AI art tools are rather small. Most AI Art codes were trained on images that were only 512px square. For this reason, as well as the fact the larger the image, the more computer resources it uses to generate, most AI images are 512px or 768px at the largest – and that’s too small for most uses, especially print.

However, thankfully, that shortcoming can be overcome by AI too – AI image upscalers. However, while there are some free ones around, the best of them (like GigaPixel AI) are commercial. So it does lessen the appeal of free AI art generators, when you have to buy software to make the images it produces usable. It also adds another step into the process of creating your images.

Secondly, is that it can take time to create the perfect image you may be wanting. AI art tools are capable of producing some absolutely stunning images, but you have to wade through a tonne of crap images to get to them, so to speak. You rarely produce perfectly usable images straight away. Instead, its a process of refining your prompts and settings, until you get something that works.

Even then, it may not be quite right. Many times, I have resorted to cobbling parts of several AI images together in my graphics software to get just what I was after – adding even more time to the process.

Thirdly, there’s some things that AI art tools are just downright terrible at! Human anatomy can trouble the AI at times. Hands and feet are particularly difficult for the AI to get right, and extra digits and/or limbs – or limbs in the wrong position – are all too common. Its incredibly frustrating to have an otherwise perfect image spoiled by body parts!

Using negative prompts can help somewhat with all this weirdness, but getting perfect looking bodily appendages is still more luck than not at the moment! Hopefully this will improve as the tech behind AI art improves.

And lastly, the controversy around the ethics of AI art – whether its ripping off artists, and taking away their livelihood – certainly can’t be overlooked. Its a big minefield, and not one I wish to ignore or downplay – so I’ll be dedicating a whole blog post to the topic. For now, I’ll just advise that if you intend to use AI for generating art – whether its for personal or business reasons – you should definitely inform yourself on both sides of the debate, and make your own decision based on your values and ethics.

Conclusion

While AI generated art is still a relatively new technology, its is advancing rapidly. There’s no doubt AI is revolutionising the way we create visuals, and solopreneurs in particular can benefit from it. There is a bit of a learning curve with it, but its highly worth it in order to create unique visuals for your business and brand, and do away with boring stock photos!

Are you a solopreneur interested in learning more about using AI for generating art? Or want to bring to life some of your image ideas, but don’t know where to start? If so, check out my 1 on 1 Interactive AI Art Sessions – where I share the knowledge I’ve gained from many hours of exploring these tools, in a fun, informative and productive way!